This project is aiming at the construction of a fusion protein consisting of a cytochrome P-450 type of enzyme, cyt. P450cam, and the two electron transporting enzymes necessary for the activity of the cyt. P-450cam. Subsequently, it is our intention to alter the substrate specificity of the enzyme by changing the active site of the cytochrome P450 part of the fusion protein by site directed mutation. The currently available structural data for P450cam (crystal structure) and putidaredoxin (NMR structure) are utilized in order to anticipate a structural conformation of a fusion protein of the three components that allows electron transfer between amino acid residues that by biochemical techniques have been shown to participate in the transfer. The genes for the three proteins are then linked together in the orientation and with amino acid linker sections that is most likely to enable this conformation. Contrary to our expectations, a PdR-Pd-P450cam fusion protein in which the C-terminus of putidaredoxin is directly linked to the N-terminus of P450cam proved to be the catalytically most effective fusion protein. This is now our vehicle of choice to transfer the activity of P450cam and mutated versions of this into other organisms, for example E. coli. New substrates to P450cam can be found with relatively good confidence by searching a database of organic compounds with the computer algorithm DOCK in conjunction with the crystal structure of P450cam to identify compounds that fit into the active site of the enzyme. We are using the reverse technique in attempts to shape the active site of the enzyme to be able to accommodate chosen organic compounds by using the DOCK3.5 algorithm to identify the amino acid residues that prevent the accommodation in wild type P450cam. These residues are then changed by site directed mutagenesis, and the validity of the predictions tested after expression of the altered enzyme in E. coli. We are currently increasing the quality of the prediction of substrates of wild type P450cam by the use of the various available crystal structures of P450cam (substrate bound, inhibitor bound, oxidized and reduced complex), and by varying the parameters used in the DOCK3.5 program. A total of 24 DOCK3.5 predicted substrates have been tested and a combination of crystal structure and DOCK3.5 parameters that will distinguish the actual substrates from nonsubstrates within this set of compounds has been defined. This optimization and the further rational engineering of mutants with changed substrate specificities relies heavily on the programs supplied by the Computer Graphics Lab.